JP2006211243A - Device and method for digital signal encoding - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable encoding of high compressibility even with respect to random signals of measured data, etc. without being restricted to a voice signal, etc. by using another reversible encoding method not using a predictor, a reversible encoding by the predictor as well. <P>SOLUTION: This device is equipped with encoders 203-1 to 203-3 which encode a residue signal calculated by a subtractor 212, and prediction coefficient specification information for specifying a prediction coefficient applied when a prediction signal used to calculate the residue signal is predicted, an encoder 203-0 for encoding digital signals of a predetermined number of samples acquired by a buffer 201 without making them pass through a predictor 202, and a code length calculator/selector 204 which selects either the encoders 203-1 to 203-3 or the encoder 203-0, on the basis of the code length of a code acquired by the encoders 203-1 to 203-3 and the encoder 203-0. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an encoding technique for lossless encoding of an input digital signal, and in particular, includes a plurality of predictors to which prediction coefficients having different code lengths are applied, and selects an optimal predictor from these predictors. The present invention relates to a digital signal encoding apparatus and a digital signal encoding method for realizing highly efficient compression of the digital signal.

In recent years, many techniques have been developed to compress the amount of data by encoding using the redundancy of digital signals such as audio signals and audio signals, and to improve the efficiency of recording / reproduction of digital signals or transmission / reception of digital signals. .
These technologies are roughly classified into lossy coding (lossy compression) and lossless coding (lossless compression).
In the former case, even if the compressed data is decoded, the data before compression cannot be completely restored, and although it is accompanied by auditory deterioration due to data compression as represented by MP3 and AAC methods, It is possible to encode a high compression rate (no significant auditory problem even at about 1/10 or less).
On the other hand, the latter is also referred to as lossless coding, and since the signal before compression (encoding) can be completely restored, auditory degradation does not occur, but a relatively low compression rate (depending on the encoded speech data). , Approximately 1/2).
Therefore, irreversible encoding is often used because of the superior compression ratio, especially in situations where the audio signal is recorded and reproduced using a portable terminal with limited memory capacity.

  However, since a program, character data, and the like have different results if even one bit is lost, lossless encoding is always used when compressing these data. Also in the audio field, some high-performance audio products cannot tolerate even the deterioration of sound quality due to lossy encoding. Furthermore, in the recent field of measuring instruments, compression technology has come to be used for measuring devices (data loggers, etc.) that measure output signals (voltage, current value, etc.) of sensors, etc. for a long time using multiple channels. In addition, the measurement data measured by such an apparatus adopts lossless encoding because data loss before and after compression is not allowed, as in the case of the program. Thus, there is a growing need for lossless encoding that can obtain exactly the same signal as before compression, and there is a strong demand for the development of a technology that improves the compression ratio, which is a technical issue. Yes.

Therefore, as a technique for improving the compression rate in lossless encoding, there is a speech encoding apparatus proposed in Patent Document 1. This device predicts speech signals input by multiple predictors with different characteristics (ie, different applied prediction coefficients), and based on the prediction results (ie, minimizes the residual signal) ) One of the plurality of predictors is selected, and an output signal (that is, a residual signal) from the selected predictor and a signal specifying the predictor (that is, a prediction coefficient) are encoded. It is configured to output later. According to the speech coding apparatus configured as described above, an optimal predictor (prediction coefficient) that can minimize the residual signal is selected according to the speech signal. The accuracy can be improved, which can contribute to the improvement of the compression rate after encoding.
However, although the speech encoding apparatus of Patent Document 1 has a plurality of predictors having different characteristics, the prediction coefficient in each predictor is premised on a fixed length (constant), and the speech signal after encoding From the viewpoint of minimizing the code length, it is difficult to say that the optimal predictor is necessarily selected. In other words, in terms of improving the prediction accuracy by the predictor, not only the coefficient of the prediction coefficient is changed, but also the order of the prediction coefficient for each predictor is changed, or the prediction method for each predictor is changed, etc. Although it is more advantageous to make the code length of the prediction coefficient for each predictor variable, in the technique of Patent Document 1, it is necessary to assume that the prediction coefficient of each predictor has a fixed length. Therefore, it was impossible to further improve the compression ratio by applying predictors with different code lengths of the prediction coefficients.

Therefore, as a technique for solving such a problem, there is a speech encoding apparatus (speech encoding method) described in Patent Document 2 previously filed by the applicant of the present application. In this apparatus, different predictive coefficients that are appropriately calculated according to frame data, for example, are assigned to each predictor instead of fixed-length predictive coefficients, and each predictor receives an audio signal input based on each predictive coefficient. The prediction is performed, and a plurality of code lengths of various patterns output from the respective predictors are compared to select a predictor that outputs the minimum code length. With such a configuration, the prediction accuracy of the audio signal is improved, and as a result, the compression rate is increased.
JP 2001-175295 A JP 2004-170494 A

However, the speech coding apparatus described in each of the above patent documents has sufficient prediction accuracy (such as speech signals, audio signals, etc.) for a signal whose sounding mechanism (speech waveform etc.) can be estimated to some extent ( Prediction performance), but unlike voice signals, audio signals, etc., for signals that cannot be estimated, such as measurement data or random signals such as random noise such as noise included in the measurement signal In some cases, the predictor cannot predict the signal sufficiently, and the compression effect as expected may not be obtained. On the other hand, for the random signal, a higher compression rate may be realized by using other lossless encoding that does not use a predictor.
Therefore, the present invention has been made in view of the above circumstances, and the object of the present invention is to use not only lossless encoding by a predictor but also other lossless encoding not using a predictor, An object of the present invention is to provide a digital signal encoding apparatus and a digital signal encoding method capable of encoding with a high compression rate not only for speech signals but also for random signals such as measurement data.

In order to achieve the above object, the present invention provides a digital signal encoding device for reversibly encoding an input digital signal, and a sample acquisition means for acquiring the input digital signal every predetermined number of samples; A plurality of prediction coefficients applied, and at least one prediction means for predicting a prediction signal of the current digital signal from a past digital signal by each prediction coefficient; a prediction signal predicted by the prediction means; and the sample acquisition means At least one residual signal calculating means for calculating a residual signal between a predetermined number of digital signals actually acquired by the method, a residual signal calculated by the residual signal calculating means, and the residual signal. Encoding prediction coefficient specifying information for specifying the prediction coefficient applied when predicting the prediction signal used to calculate the difference signal At least one first encoding means, second encoding means for encoding a digital signal of a predetermined number of samples acquired by the sample acquisition means without going through the prediction means, and the first encoding Coding selection means for selecting one of the first coding means and the second coding means based on the code length of the code obtained by the means and the second coding means. Thus, the digital signal encoding device is configured.
With this configuration, even a digital signal such as a random signal for which a sufficient compression effect cannot be obtained by encoding using a predictor (prediction means) can be encoded at a high compression rate. It becomes.
Further, it further comprises effective bit number detecting means for detecting the effective bit number of the digital signal having the predetermined number of samples acquired by the sample acquiring means, and the second encoding means is detected by the effective bit number detecting means. It is desirable to encode the effective bit number and the effective bit number specifying information for specifying the effective bit number. For example, if the effective bit range in the frame data is known, the other bit ranges are not necessary as information, so that the frame data can be compressed into data for the effective bit range. Of course, such encoding (compression) is an example of the second encoding means, and other various encodings can be applied as long as the encoding does not use a predictor.

As an example of the prediction coefficient in the prediction means, a prediction coefficient calculated by linear prediction analysis from a predetermined number of samples of the speech signal acquired by the sample acquisition means can be considered.
Another example of the prediction coefficient in the prediction means is a prediction coefficient obtained by vector quantization based on a prediction coefficient calculated by linear prediction analysis from the speech signal of the predetermined number of samples acquired by the sample acquisition means. Conceivable.
As another example of the prediction coefficient in the prediction means, the prediction coefficient used in the first encoding means selected in the past by the encoding selection means can be considered.

Note that the present invention is a digital signal encoding method for reversibly encoding an input digital signal by grasping it as a speech encoding method applied to the speech encoding apparatus having the above-described configuration. A sample acquisition step for acquiring a signal every predetermined number of samples, and at least one prediction step in which a plurality of prediction coefficients are applied, and a prediction signal of the current digital signal is predicted from a past digital signal by each prediction coefficient; At least one residual signal calculation step for calculating a residual signal between the prediction signal predicted by the prediction step and the current digital signal of a predetermined number of samples actually acquired by the sample acquisition step; Before applying the residual signal calculated by the difference signal calculating step and the prediction signal used for calculating the residual signal Encode at least one first encoding step for encoding prediction coefficient specifying information for specifying a prediction coefficient, and a digital signal of a predetermined number of samples acquired by the sample acquisition step without going through the prediction step. Based on the second encoding step and the code length of the code obtained by the first encoding step and the second encoding step, the first encoding step and the second encoding step It can also be considered as a digital signal encoding method characterized by comprising an encoding selection step of selecting any one.
In this case, the method further comprises an effective bit number detection step for detecting the effective bit number of the digital signal of the predetermined number of samples acquired by the sample acquisition step, and the second encoding step is detected by the effective bit number detection step. It is preferable that the effective bit number and the effective bit number specifying information for specifying the effective bit number are encoded.
Since the operation in this case is the same as that described above, the description thereof is omitted here.

  As described above, according to the present invention, the first encoding means for encoding using the predictor and the second encoding means for encoding without using the predictor are provided. Even a digital signal such as a random signal for which a sufficient compression effect cannot be obtained by encoding using a predictor (prediction unit) is encoded by the second encoding unit (encoding efficiency (compression)). Rate).

Hereinafter, embodiments and examples of the present invention will be described with reference to the accompanying drawings so that the present invention can be understood. It should be noted that the following embodiments and examples are examples embodying the present invention and do not limit the technical scope of the present invention.
FIG. 1 is a block diagram showing a schematic configuration of a digital signal encoding apparatus A according to the first embodiment of the present invention. FIG. 2 is a block diagram of the digital signal encoding apparatus B according to the second embodiment of the present invention. FIG. 3 is a block diagram showing a schematic configuration, and FIG. 3 is a block diagram showing a schematic configuration of a digital signal decoding apparatus C for decoding the bitstream E1 generated by the digital signal encoding apparatus A according to the first embodiment of the present invention. FIG. 4 is a block diagram showing a schematic configuration of a digital signal decoding apparatus D that decodes the bit stream E2 generated by the digital signal encoding apparatus B according to the second embodiment of the present invention, and FIG. FIG. 6 is a diagram illustrating an example of a digital signal used for explaining the encoding of the embodiment, FIG. 6 is a diagram illustrating a code length of a predictor identifier and prediction coefficient specifying information in the bit stream, and FIG. FIG. 8 schematically shows another example of the bit stream configuration, FIG. 8 schematically shows another example of the bit stream configuration, and FIG. 9 shows the average bit when Golomb coding is applied to a predetermined random signal. FIG. 10 is a diagram showing an example of the digital signal, FIG. 11 is a diagram showing the appearance frequency for each amplitude in the speech signal, and FIG. 12 is a diagram showing the appearance frequency for each amplitude in the prediction residual signal. .

<< First Embodiment >>
Here, the digital signal encoding apparatus A (hereinafter abbreviated as “encoding apparatus A”) according to the first embodiment of the present invention is embodied as shown in the block diagram of FIG.
As shown in the figure, the encoding apparatus A according to the present embodiment is an apparatus that performs lossless encoding of an input digital signal, and includes a buffer 201 (an example of a sample acquisition unit) and predictors 202-1 to 202-. 3 (an example of prediction means), subtractors 212-1 to 212-3 (an example of residual signal calculation means), an encoder 203-0 (an example of a second encoding means), and encoders 203-1 to 203 -3 (an example of a first encoding unit), a code length calculator / selector 204 (an example of an encoding selection unit), a bit stream generator 205, and a memory unit 213.
The PCM data (that is, a digital signal such as a voice signal or a measurement signal) input to the encoding device A is encoded (compressed) below and is output as encoded data (compressed data). The processing will be described along the flow of processing.

(Buffer 201)
First, the PCM data input to the encoding device A is input from the input terminal 200 to the buffer 201.
In the buffer 201, the PCM data is buffered so that a certain number of samples (for example, 800 samples) can be processed in units called frames.
FIG. 10 schematically shows how PCM data input to the buffer 201 is sequentially buffered while being framed. As shown in the figure, the input PCM data F is framed every predetermined number of samples and sequentially buffered in the buffer 201.
When there are a plurality of PCM data channels, the buffer 201 may perform processing for removing the correlation between the channels. For example, if the PCM data is two channels of Lch and Rch, the calculation processing of Lch + Rch or Lch−Rch corresponds to this.
Next, frame-unit PCM data (hereinafter simply referred to as “frame data”) buffered by the buffer 201 is input to the encoder 203-0 and the predictors 202-1 to 202-3, respectively. .
Conventionally, the frame data is input only to the predictor 202. However, in the present invention, not only each predictor 202 but also the frame data is directly input to the encoder 203-0. At 0, the frame data is encoded without going through the predictor 202. As described above, the present invention is different from the conventional one in that a path for directly inputting the frame data to the encoder 203-0 is provided without using the predictor 202. For example, when the frame data is a signal having randomness (random signal), the compression rate (encoding efficiency) is higher when encoding without using a predictor. This is because the possibility is considered.

が予測される。

但し，αｎ（ｎ＝１，２，・・，Ｎ）は予測係数である。
ここで，該予測器２０２には，各予測器に対し，異なる符号長を有する予測係数が割り当てられ，その予測精度の向上が図られている。
そのため，当該符号化装置Ａでは，後述するように，前記予測器２０２のうち一つの予測器を選択するに当たり，その予測器により予測された予測信号を用いて算出される残差信号だけでなく，その予測信号を予測する際に適用された「予測係数を特定する情報（以下「予測係数特定情報」と称す）」の符号長を考慮する必要がある。
ここでは，先ず，前記予測器２０２の夫々に適用される予測係数，及び予測係数特定情報の符号長について，以下に説明する。
（予測器２０２−１）
先ず，前記予測器２０２−１には，バッファリングされたフレームデータを対象として線形予測分析器２０７で算出された線形予測係数２０９を割り当てる。
これにより，フレームデータに応じて適宜算出される予測係数によって予測を行うことが可能となり，一定の予測係数による予測に較べ，予測精度を著しく向上させ得る。
尚，前記線形予測分析器２０７における線形予測係数２０９の算出方法に関する詳細は，「音のコミュニケーション工学」（社団法人日本音響学会編，コロナ社，１９９６／８／３０初版）に詳しいため，ここでは省略する。
このように前記線形予測係数２０９が適用される当該予測器２０２−１では，前記予測係数特定情報の符号長を，前記線形予測分析器２０７で算出される前記線形予測係数２０９が４次，各次数毎の予測係数が８ｂｉｔの場合を考えると，４×８ｂｉｔ＝３２ｂｉｔに，後述する予測器識別子の２ｂｉｔを加算した３４ｂｉｔとすることが可能である（図６（ｄ）参照）。無論，この符号長は一例であって，予測係数の次数や各次数毎の符号長は前述の値に限定されるものではない。
（予測器２０２−２）
そして，前記予測器２０２−２には，１フレーム前で使用した予測係数２１１を割り当てる。つまり，前回，前記符号長計算／選択器２０４により選択された符号データの符号化の際に用いられた予測係数を一旦メモリ部２１３に保存しておき，現在のフレームデータを符号化する際に，その保存された前記予測係数２１１を前記メモリ部２１３から読み出して使用する。
これは，現在のフレームデータと前回のフレームデータとの間には何らかの相関性があるとの仮定に基づくものであり，例えば，同一或いは類似したデータが繰り返されるＰＣＭデータに対して有効であると解される。
このように前記予測係数２１１が適用される当該予測器２０２−２では，前記予測係数特定情報の符号長を，過去の前記予測係数２１１を使用したことを表すインデックス（フラグ）のみとすることが可能である。つまり，前記フラグの「０」「１」に応じ，前記予測係数２１１の使用／不使用を判定するよう構成することができる。なお，本実施形態では，識別対象が多いため，前記予測係数２１１の使用／不使用の判定は，２ｂｉｔで表された予測器識別子（図６（ｂ）参照）を用いるが，もちろん，この予測器識別子は，識別対象数に応じた任意のビット数（ビット幅）とすることも可能である。尚，本実施形態では，簡単のため過去１フレームのみの予測係数を使用する形態としているが，更に過去のフレームに使用された予測係数を保存しておき，それらを使用する形態とすることも可能である。
（予測器２０２−３）
そして，前記予測器２０２−３には，複数の予測係数が予め記憶された予測係数テーブル２１０から選択された予測係数を割り当てる。
ここで，前記予測係数テーブル２１０から一つの予測係数を選択する手法として，本実施形態では，前記線形予測分析器２０７で算出された前記線形予測係数２０９に最も近い予測係数２０９′を選択し，該予測係数２０９′を特定する前記予測係数テーブル２１０の行番号（以下，テーブルインデックスといい，図１参照）を符号化する，いわゆるベクトル量子化法を用いる。
このような構成により，予め記憶された複数の予測係数から最も高い予測精度を発揮し得る予測係数を選択することが可能となり，より高い圧縮率での符号化を実現し得る。
このように前記予測係数２０９′が適用される当該予測器２０２−３では，前記予測係数特定情報の符号長を，該予測係数２０９′毎に設定された前記テーブルインデックスに，後述する予測器識別子の２ｂｉｔを加算したｂｉｔ数とすることが可能である。ここで，前記予測係数テーブル２１０のテーブルが２５６通りの場合を考えると，その符号長は，８ｂｉｔとなるため，この場合の予測係数特定情報は１０ｂｉｔとなる（図６（ｃ）参照）。無論，このテーブル数は一例であって，この値に特定されるものではない。
更に，前記予測係数テーブル２１０に記憶されている前記予測係数２０９′は，上述した線形予測係数に限定されるものではなく，これと等価なＰＡＲＣＯＲ係数，或いはＬＳＰ（Ｌｉｎｅ Ｓｐｅｃｔｒｕｍ Ｐａｉｒ）係数等を用いることも可能である。 (Predictor 202)
First, a path through which frame data is input to the predictor 202 will be described.
In the predictor 202, linear prediction processing is performed based on the predetermined prediction coefficient αn assigned to each as shown in the following equation (1), and the predicted value of the current sample from the past sample data x (i).

Is predicted.

Here, αn (n = 1, 2,..., N) is a prediction coefficient.
Here, to the predictor 202, prediction coefficients having different code lengths are assigned to the respective predictors, and the prediction accuracy is improved.
Therefore, in the encoding apparatus A, when selecting one predictor from the predictors 202 as described later, not only the residual signal calculated using the prediction signal predicted by the predictor but also the predictor 202 is selected. Therefore, it is necessary to consider the code length of “information for specifying a prediction coefficient (hereinafter referred to as“ prediction coefficient specifying information ”)” applied when predicting the prediction signal.
Here, first, the prediction coefficient applied to each of the predictors 202 and the code length of the prediction coefficient specifying information will be described below.
(Predictor 202-1)
First, a linear prediction coefficient 209 calculated by the linear prediction analyzer 207 with respect to the buffered frame data is assigned to the predictor 202-1.
As a result, it is possible to perform prediction using a prediction coefficient that is appropriately calculated according to the frame data, and the prediction accuracy can be significantly improved as compared with prediction using a constant prediction coefficient.
The details of the calculation method of the linear prediction coefficient 209 in the linear prediction analyzer 207 are detailed in “Sound Communication Engineering” (edited by the Acoustical Society of Japan, Corona, 1996/8/30, first edition). Omitted.
As described above, in the predictor 202-1 to which the linear prediction coefficient 209 is applied, the code length of the prediction coefficient specifying information is the fourth order of the linear prediction coefficient 209 calculated by the linear prediction analyzer 207. Considering the case where the prediction coefficient for each order is 8 bits, it is possible to obtain 34 bits obtained by adding 2 bits of a predictor identifier described later to 4 × 8 bits = 32 bits (see FIG. 6D). Of course, this code length is an example, and the order of the prediction coefficient and the code length for each order are not limited to the above-mentioned values.
(Predictor 202-2)
The predictor 202-2 is assigned the prediction coefficient 211 used one frame before. That is, the prediction coefficient used in the previous encoding of the code data selected by the code length calculator / selector 204 is temporarily stored in the memory unit 213, and the current frame data is encoded. The stored prediction coefficient 211 is read from the memory unit 213 and used.
This is based on the assumption that there is some correlation between the current frame data and the previous frame data. For example, it is effective for PCM data in which the same or similar data is repeated. It is understood.
In this way, in the predictor 202-2 to which the prediction coefficient 211 is applied, the code length of the prediction coefficient specifying information may be only an index (flag) indicating that the past prediction coefficient 211 has been used. Is possible. That is, the use / non-use of the prediction coefficient 211 can be determined according to “0” and “1” of the flag. In this embodiment, since there are many identification targets, the use / non-use determination of the prediction coefficient 211 uses a predictor identifier represented by 2 bits (see FIG. 6B). The device identifier can be an arbitrary number of bits (bit width) according to the number of identification objects. In this embodiment, for the sake of simplicity, the prediction coefficient of only one past frame is used. However, the prediction coefficients used in the past frame may be stored and used. Is possible.
(Predictor 202-3)
The predictor 202-3 is assigned a prediction coefficient selected from the prediction coefficient table 210 in which a plurality of prediction coefficients are stored in advance.
Here, as a method for selecting one prediction coefficient from the prediction coefficient table 210, in the present embodiment, the prediction coefficient 209 ′ closest to the linear prediction coefficient 209 calculated by the linear prediction analyzer 207 is selected, A so-called vector quantization method is used which encodes the row number (hereinafter referred to as a table index, see FIG. 1) of the prediction coefficient table 210 for specifying the prediction coefficient 209 ′.
With such a configuration, it is possible to select a prediction coefficient that can exhibit the highest prediction accuracy from a plurality of prediction coefficients stored in advance, and encoding at a higher compression rate can be realized.
As described above, in the predictor 202-3 to which the prediction coefficient 209 ′ is applied, the code length of the prediction coefficient specifying information is set in the table index set for each prediction coefficient 209 ′, as will be described later. It is possible to obtain the number of bits obtained by adding 2 bits. Here, considering the 256 cases of the prediction coefficient table 210, the code length is 8 bits, so the prediction coefficient specifying information in this case is 10 bits (see FIG. 6C). Of course, this number of tables is an example and is not specified by this value.
Further, the prediction coefficient 209 ′ stored in the prediction coefficient table 210 is not limited to the linear prediction coefficient described above, and a PARCOR coefficient equivalent to this, a LSP (Line Spectrum Pair) coefficient, or the like is used. It is also possible.

Next, the code length of the prediction coefficient specifying information in the prediction coefficient applied to each of the predictors 202 will be described in detail with reference to FIG.
This prediction coefficient specifying information is composed of a predictor identifier and other data portions as shown in FIG.
The predictor identifier is arranged at the head portion of the prediction coefficient specifying information, and the prediction coefficient is applied to which predictor among the predictors 202 described above according to the prediction coefficient specifying information. It is a flag indicating whether it is a coefficient.
Specifically, as shown in FIG. 6B, if the first 1 bit is 1 and the 2nd bit is 0, it is the prediction coefficient of the previous frame (that is, the predictor 202-2 is applied). 6 (c), if the first bit is 0 and the second bit is 1, the prediction coefficient 209 ′ selected from the prediction coefficient table 210 (that is, The predictor 202-3 is applied), and if the first bit is 0 and the second bit is 0 as shown in FIG. 6 (d), the linear prediction analyzer 207 calculates the linear It indicates that the prediction coefficient is 209 (that is, the predictor 202-1 is applied).
On the other hand, the data portion is arranged after the predictor identifier, and information necessary for specifying the prediction coefficient is given by the data portion.
Specifically, as shown in FIG. 6B, the prediction coefficient of the previous frame is uniquely identified only by the predictor identifier, so that new information (data portion) is not necessary, but FIG. The prediction coefficient 209 ′ selected from the prediction coefficient table 210 corresponds to the table index of the prediction coefficient 209 ′, and is calculated by the linear prediction analyzer 207 as shown in FIG. For the linear prediction coefficient 209, the linear prediction coefficient 209 itself corresponds to this.
Thus, in the encoding apparatus A according to the present embodiment, the predictor 202-1 is selected as the code length of the prediction coefficient specifying information applied to the predictor 202 (see FIG. 6D). ) In the case of 2 + 32 = 34 bits and the predictor 202-2 is selected (see FIG. 6B), and in the case of 2 bits and the predictor 202-3 is selected (see FIG. 6C). 2 + 8 = 10 bits, which is understood to be different for each predictor.
Therefore, in the encoding apparatus A according to the present embodiment, the code length calculator / selector 204 described later takes into account the difference in code length of the “information for specifying the prediction coefficient”, and then the plurality of the predictors. From 202, an optimal predictor (that is, an encoder that can perform encoding with the highest compression rate) is selected.
In the above description, the predictor identifier has a fixed length of 2 bits. Of course, the predictor identifier may have a variable length.

と，前記バッファ２０１から入力されたサンプルデータｘ（ｉ）とを用い，下式（２）の如く，残差信号ｄ（ｉ）が算出される。

(Subtractor 212)
As described above, each of the predetermined prediction coefficients is applied, and the predictor 202 that has predicted the prediction signal of the current sampling data from the past sampling data outputs the prediction signal to the subtracters 212-1 to 212-3. .
In the subtractor 212, the prediction signal input from the predictor 202.

Then, using the sample data x (i) input from the buffer 201, a residual signal d (i) is calculated as shown in the following equation (2).

(Encoder 203-0)
Of the encoders 203-0 to 203-3, the frame data buffered in the buffer 201 is directly input to the encoder 203-0 without going through the predictor 202, as described above. The At this time, for example, by adding a predictor identifier (see FIG. 6A) indicating that a prediction coefficient is not used for encoding to the head of the input frame data in the buffer 201, the frame data is The encoder 203-0 can be input. Specifically, as shown in FIG. 6A, if the first 1 bit is 1 and the second bit is 1, the frame data is directly input to the encoder 203-0 (that is, the prediction coefficient is not used). )
The frame data input to the encoder 203-0 is encoded by the encoder 203-0, and then the encoded frame data (hereinafter referred to as “encoded frame”) is converted to the code length calculator / selector. It is output to 204.
The method of encoding the frame data performed by the encoder 203-0 is not particularly limited. For example, various encoding methods such as entropy encoding described later can be applied. is there.

(Encoders 203-1 to 203-3)
On the other hand, the encoders 203-1 to 203-3 correspond to the residual signal d (i) for each predictor 202 and the residual signal d (i) (that is, the residual signal d ( The prediction coefficient specifying information (applied when predicting the prediction signal used for the calculation of i) is input. In the encoders 203-1 to 203-3, the residual signal d (i) acquired from the subtractor 212 and the prediction coefficient specifying information acquired from the predictor 202 are encoded, and then encoded. The residual signal and the prediction coefficient specifying information are output to the code length calculator / selector 204. Hereinafter, the encoded residual signal and prediction coefficient specifying information and the encoded frame are collectively referred to as “encoding candidates”.
First, with respect to the residual signal d (i), it is desirable to perform an encoding method (so-called entropy encoding) using the deviation in appearance frequency.
The advantage of encoding the residual signal in the encoders 203-1 to 203-3 is that entropy (p [i] × log2 (1 / p, where the appearance probability of the amplitude k [i] is p [i]). The sum of [i]) can be reduced.
In this regard, FIG. 11 showing the cumulative frequency of amplitude of certain PCM data (16 bits), and a predicted value of the PCM data using a predetermined prediction coefficient, and a residual signal calculated using the predicted value. This will be described with reference to FIG.
As apparent from both figures, by calculating the residual signal of the PCM data, the amplitude distribution for the signal is concentrated in the vicinity of zero. Therefore, the compression rate can be improved by encoding using the bias (concentration) of the amplitude distribution.
Specifically, it is possible to increase the compression efficiency by assigning a short bit length code to a signal with an amplitude close to 0 and assigning a long bit length code as the amplitude increases. . Although the above description is a conceptual explanation, a technique such as a Huffman coding method, an arithmetic coding method, or a range coder, which is a well-known technique, is used as a coding method that more efficiently embodies these concepts. It is possible.
In this way, encoding with higher compression efficiency is possible by encoding the residual signal in the encoders 203-1 to 203-3.
On the other hand, the prediction coefficient specifying information is not particularly limited with respect to the encoding method, and the structures shown in FIGS. 6B to 6D can be regarded as encoding, and FIGS. The prediction coefficient specifying information having the structure shown in (d) may be further encoded using the above-described entropy encoding or the like.
That is, the encoders 203-1 to 203-3 each encode the residual signal d (i) acquired from the subtractor 212 and the prediction coefficient specifying information acquired from the predictor 202 (both of them). The encoding method may be the same or different), and has a function of outputting the encoding candidates formed by combining the encoded codes to the code length calculator / selector 204.

(Code length calculator / selector 204)
In the code length calculator / selector 204, when selecting one of a plurality of encoding candidates input from the encoder 203 (203-0 to 203-3), the code length of the encoding candidate is the smallest. A certain one is selected, and the selected encoding candidate is output to the bit stream generator 205.
That is, the code length calculator / selector 204 includes not only encoded data composed of the residual signal encoded by the encoders 203-1 to 203-3 and the prediction coefficient specifying information, but also an encoder. It is configured to select the one with the shortest code length from all the encoding candidates including the frame data (encoded frame) encoded without a predictor in 203-0. Therefore, according to the present embodiment, a digital signal that is encoded at a high compression rate in the encoding using the predictor 202 and that can be encoded at a high compression rate according to another encoding method. Even when input, an appropriate encoder capable of encoding with high compression rate can be selected for the digital signal.
Since the selection by the code length calculator / selector 204 is performed in units of frame data, the bit stream E1 (FIG. 7) described later includes the data encoded by the encoder 203-0 and the encoder 203-. Data encoded in 1-203-3 will be mixed.

(Bitstream generator 205)
Subsequently, one encoding candidate selected by the code length calculator / selector 204 is packed as a bit stream by the bit stream generator 205 and then output from the output terminal 206.
An example of the bit stream packed by the bit stream generator 205 is shown in FIG. The bit stream E1 shown in the figure shows an example of a bit stream generated from PCM data with one channel.
As shown in the figure, the bit stream E1 is roughly divided into a file header 608 and subsequent frame data. Further, the frame data is divided into frame data (1) 609, frame data (2) 610,. Here, the frame data (1) 609 includes frame data (encoded frame) encoded by the encoder 203-0 without passing through the predictor 202, and the frame data (2) 610 includes any prediction data. It is assumed that residual information and the like encoded by any of the encoders 203-1 to 203-3 are included.
First, the components of the file header 608 will be described.
The file header 608 is further divided into a header 601 and a head sample value 602. When the bit stream E1 is a bit stream generated from PCM data having two channels (for example, a stereo composed of Lch and Rch), the file header 608 includes a head sample value for each channel. included.
The header 601 includes, for example, information governing the entire bit stream (PCM data), for example, information such as sampling rate, number of channels, average bit rate, and the like.
Also, the first sample value 602 stores the first sample value of a predetermined channel. Note that the first sample value 602 is necessary to restore the first frame to be decoded, and at least the order (ie, n) of the prediction coefficient αn may be secured for each channel. Alternatively, if the initial values of past samples necessary for linear prediction are set to 0 for both the encoder and the decoder, the leading sample value 602 may be omitted. Thus, the header 601 and the head sample value 602 constituting the file header 608 have a fixed length.

Next, components of the frame data (1) 609 will be described.
The frame data (1) 609 is divided into system information 603 for identifying an encoding method (system) and an encoded frame 604.
Needless to say, the encoded frame 604 is selected by the code length calculator / selector 204 as an encoding candidate input from the encoder 203-0 (that is, frame data encoded by the encoder 203-0). In that case. The scheme information 603 is information for identifying the coding method (method) of the coded frame 604. Specifically, the coding method applied to the coding in the encoder 203-0 is selected. It is information to identify. The method information 603 is used when the bit stream E1 is decoded.

Next, components of the frame data (2) 610 will be described.
The frame data (2) 610 is further divided into method information 605, prediction coefficient specifying information 606, and residual signal information 607.
The method information 605 is information for identifying an encoding method (method) obtained by encoding the prediction coefficient specifying information 606 and the residual signal information 607, and specifically, an encoding method in a corresponding encoder. , Used predictor 202, used prediction coefficient, and the like. This method information 605 is used when the bit stream E1 is decoded.
As described above, the prediction coefficient specifying information 606 is information for specifying a prediction coefficient, and is used when the bit stream E1 is decoded. A device (digital signal decoding device C) for decoding PCM data from the bitstream E1 based on the prediction coefficient specifying information 606 will be described later.
The residual signal information 607 is a code obtained by encoding the residual signal. Thus, the prediction coefficient specifying information 606 and the residual signal information 607 constituting the frame data (2) 610 are both variable length.
Needless to say, the arrangement order of each information or the number of channels is not limited to the example shown in FIG.

Next, an example of a digital signal decoding apparatus C that can acquire and decode the bit stream E1 (see FIG. 7) generated by the encoding apparatus A will be described with reference to FIG.
Hereinafter, the processing of each unit until the bit stream E1 input to the digital signal decoding device C (hereinafter referred to as “decoding device C”) shown in FIG. 3 is decoded and output as PCM data will be described. A description will be given along the flow. The components having the same names as the components of the encoding device A shown in the block diagram of FIG. 1 have the same configuration. Therefore, the description of the encoding device A is referred to above, and a detailed description thereof is given here. Is omitted.
First, the bit stream E1 input to the speech decoding apparatus C is input to the bit stream decoder 502 via the input terminal 501.
The bit stream decoder 502 decodes the segment of the bit stream E1 and decomposes the bit stream E1 into signal components, and then distributes the signal components.
Specifically, in the bitstream decoder 502, the file header 608 of the bitstream E1 is decomposed into a header 601 and a head sample value 602, and the head sample value 602 is distributed to the buffer 504 and the predictor 503, and PCM data Is output as a part (first PCM data).

  Also, the frame data (1) 609 included in the bit stream E1 is divided into the scheme information 603 and the encoded frame 604, and each data is output to the predictor 503. However, since the frame data (1) 609 is data encoded by the encoder 203-0 without using the predictor 202, the predictor 503 uses the prediction added to the head of the frame data (1) 609. The device identifier (see FIG. 6 (a)) is determined, the prediction coefficient is processed as “0”, decoded by a decoding method corresponding to the encoding method specified by the method information 603, and then decoded. PCM data is sequentially transferred to the buffer 504. Of course, without outputting the frame data (1) 609 to the predictor 503, a decoder (not shown) corresponding to the encoder 203-0 is provided in another path, and the frame data (1) 609 is output to the decoder. Then, the decryption may be performed.

On the other hand, the frame data (2) 610 is divided into method information 605, prediction coefficient specifying information 606, and residual signal information 607. The method information 605 and the residual signal information 607 are output to the predictor 503, and the prediction data The coefficient specifying information 606 is output to the prediction coefficient forming unit 505.
The prediction coefficient specifying information 606 is distributed to the prediction coefficient forming unit 505 and used to specify a prediction coefficient to be applied to the predictor 503. Specifically, when encoding the PCM data using the encoding apparatus A according to the present embodiment, the prediction coefficient applied to the predictor 202 (see FIG. 1) is specified. That is, based on the predictor identifier included in the prediction coefficient specifying information 606 and a data portion accompanying the predictor identifier (see FIG. 6), the prediction coefficient 506 of the previous frame stored in the memory unit 509 (the predictor). The prediction coefficient 211 applied to 202-2), the prediction coefficient stored in the prediction coefficient table 507 in advance (corresponding to the prediction coefficient 209 ′ applied to the predictor 202-3), or the prediction coefficient specifying information. One of the linear prediction coefficients 209 (corresponding to the prediction coefficient 209 applied to the predictor 202-1) transmitted as the data portion 606 is selected. The prediction coefficient specifying information 606 is acquired for each frame data encoded using the predictor 202, and a prediction coefficient applied to the encoding (prediction) is appropriately selected for each frame data. Is done. By selecting a prediction coefficient to be applied to the predictor 503 by this procedure, the prediction coefficient is the same as the prediction coefficient applied to the predictor 202 (see FIG. 1) when the PCM data is encoded. It becomes possible to do.
The output signal of the predictor 503 and the residual signal information 607 are added to be decoded into PCM data, and the decoded PCM data is sequentially transferred to the buffer 504.

The PCM data transferred in this way and stored in the buffer 504 is output as an output signal of the buffer 504 via the output terminal 508, thereby completing the decoding into PCM data.
Thus, in the decoding apparatus C, it is possible to perform lossless decoding of the original PCM data according to the reverse procedure of the encoding apparatus A.
The decoding apparatus C has a configuration in which the buffer 504 is provided so as to correspond to the buffer 201 in the encoding apparatus A described with reference to FIG. 1. However, the buffer 504 is not an indispensable component. The decoded PCM data may be sequentially output via the output terminal 508 without being sequentially buffered by the buffer 504. However, as described in the above-described encoding procedure, the buffer 201 in the encoding apparatus A performs processing using the correlation between the channels (for example, Lch + Rch or Lch-Rch in a 2ch audio signal). If so, it is necessary to perform reverse processing in the buffer 504.

<< Second Embodiment >>
Next, a digital signal encoding device B (hereinafter abbreviated as “encoding device B”) according to a second embodiment of the present invention will be described with reference to the block diagram of FIG. In FIG. 2, the same components as those of the encoding device A of the first embodiment described above are denoted by the same reference numerals, and the description thereof is omitted.
In the encoding apparatus A according to the first embodiment shown in FIG. 1 described above, it has been described that the encoder 203-0 can apply various encoding methods such as entropy encoding.
However, when the randomness of the input digital signal is strong, that is, when the input digital signal is a random signal, there is little bias in the cumulative frequency of amplitude (see FIGS. 11 and 12), so entropy coding is performed. Is not executed effectively, that is, the coding rate (compression rate) is lowered. An example of the random signal is measurement data measured by a measuring device such as a data logger. Since such measurement data includes random noise, the randomness of the data tends to be strong.
In general, a digital signal (digital data) has predetermined information represented by all the bits of the data length (for example, 16 bits). However, in a random signal typified by the measurement data, substantially all In some cases, information is not represented by the number of bits, but only by a predetermined number of bits (bit range, bit width) from the beginning of the data. In general, such a bit number is called an effective bit number.
It is known that the number of effective bits is a relatively small value for the random signal.
Here, in FIG. 9, an average bit length (average code length) when a known Golomb encoding (Golomb-Rice encoding), which is one of entropy encoding, is performed on a random signal having a 16-bit width. Indicates. Specifically, the lowermost column of the table shown in FIG. 9 shows an average bit length (average code length) when a random signal in which integer values of −4 to 3 are randomly generated is Golomb encoded.
As described above, even if any coding method of Goram coding 1 to 4 is applied to a random signal in which integer values of −4 to 3 are randomly generated, the average bit length after the coding is 3. 5 to 4.5. However, since the integer value of -4 to 3 indicated by the random signal can be represented by 3 bits, all the bit strings other than the leading 3 bits are removed from the 16-bit width, and the 16-bit random signal is converted to 3 It should be possible to encode into bit data.

  Therefore, in the encoding device B according to the second embodiment of the present invention, as shown in the block diagram of FIG. 2, the number of effective bits of the frame data input from the buffer 201 to the encoder 203-0 is detected. A bit range detector 302 (an example of effective bit number detection means) is provided, and information for specifying the effective bit number detected by the bit range detector 302 (hereinafter referred to as “effective bit number specifying information”) is the encoder 203. The encoder 203-0 is configured to encode frame data having a predetermined bit width (for example, 16-bit width) into data having the effective number of bits. With this configuration, it is possible to further increase the coding rate (compression rate) of highly random data.

Here, a specific example of encoding using the number of effective bits will be described below with reference to FIG. 2, FIG. 5 and FIG.
For the sake of simplicity, as shown in FIG. 5, 16 pieces of 16-bit sample data (samples 401 to 416) representing any integer value of −4 to 3 in decimal notation are assembled. Consider the frame data 40. The frame data 40 is a data group having a time axis in the vertical direction. The frame data 40 represents the absolute value of an integer value of −4 to 3 with the two bits (bit numbers 0 and 1) from the beginning, and the third bit (bit number 2) from the beginning represents the sign. To express. That is, the third bit from the head is a valid bit, and the number of valid bits is three. Due to the nature of the sample data, the other bits except the valid bits have the same bit value as the sign.
First, when the frame data 40 is obtained by being framed and buffered by the buffer 201, the frame data 40 is then input to the encoder 203-0. At this time, the number of effective bits of the frame data 40 is detected by the bit range detector 302 provided in the input path of the frame data 40. Such detection can be performed by, for example, determining the valid bit range based on whether or not the value of each bit number of the sample data appears at random. Of course, it is not limited to such a method.
When the effective bit number is detected, effective bit number specifying information for specifying the effective bit number is input from the bit range detector 302 to the encoder 203-0.
Next, the encoder 203-0 encodes the frame data 40 based on the input frame data 40 and effective bit number information. Specifically, 16-bit sample data 401 to 416 are encoded into 16 sample data corresponding to the number of effective bits (hereinafter referred to as “bit limit information”, range 43 in FIG. 5). The bit limit information 43 is compressed into code data 44 obtained by adding the effective bit number information (bit width 42 in FIG. 5).

An example of the bit stream packed by the bit stream generator 205 of the encoding device B is shown in FIG. The bit stream E2 shown in the figure shows an example of a bit stream generated from PCM data with one channel.
As shown in the figure, the bit stream E2 is roughly divided into a file header 608 and subsequent frame data. Further, the frame data is divided into frame data (1) 609, frame data (2) 610,. Further, the frame data (1) 609 is divided into method information 603 for identifying an encoding method (method) and the code data 44 (see FIG. 5). Note that the same data (information) included in the bit stream E1 is denoted by the same reference numeral and description thereof is omitted.
Thus, by encoding using the effective number of bits, before encoding, 256 bits (16 × 16 bits) of frame data 40 is converted into 51 bits (3 + 3 × 16 bits) of code data, that is, It is encoded into code data 44 compressed to about 20%.

Next, an example of a digital signal decoding apparatus D that can acquire and decode the bit stream E2 (see FIG. 8) generated by the encoding apparatus B will be described with reference to FIG. Hereinafter, the processing of each unit until the bit stream E2 input to the digital signal decoding device D (hereinafter referred to as “decoding device D”) shown in FIG. 4 is decoded and output as PCM data will be described. A description will be given along the flow. Note that the same components as those of the decoding device C are denoted by the same reference numerals and detailed description thereof is omitted.
The decoding device D according to the second embodiment has substantially the same configuration as the decoding device C, but further includes a bit range extender 510 and a path switch 511 as shown in FIG. The configuration is different.
First, the bit stream E2 input to the speech decoding apparatus D is input to the bit stream decoder 502 via the input terminal 501, and after the segments of the bit stream E2 are decoded and decomposed for each signal component. The signal components are sorted.
Specifically, the file header 608 is decomposed into a header 601 and a head sample value 602, the head sample value 602 is distributed to the buffer 504 and the predictor 503, and is output as a part of PCM data (head PCM data). .
Frame data (1) 609 included in the bit stream E2 is divided into method information 603, bit number specifying information 42, and bit restriction information 43, and frame data (2) 610 includes method information 605, prediction coefficient specifying information 606, It is divided into residual signal information 607.
Among the divided data, system information 603, 605, bit restriction information 43, and residual signal information 607 are output to the path switch 511. Meanwhile, the bit number specifying information 42 is output to the bit range extender 510.
The path switch 511 outputs the information output from the bit stream decoder 502 to either the predictor 503 or the bit range extender 510 in order to output the input method information 603 and the input frame. The route R1 and the route R2 are switched with reference to data and the like. Specifically, it is determined whether the input data includes the bit limit information 43 or the residual signal information 607. If the input data includes the bit limit information 43, the input data is switched to the route R2, and the remaining data is included. When the difference signal information 607 is included, the route R1 is switched. In this way, data is distributed to each part. For the processing performed by the predictor 503, refer to the description of the decoding device C.
When the bit limit expander 510 outputs the bit limit information 43 to the bit range extender 510, the bit range extender 510 determines the limit of the input bit limit information 43 based on the bit number specifying information 42 corresponding to the bit limit information 43. A process of extending (decoding) the original frame data is executed. For example, as shown in FIG. 5, processing is performed to obtain bit number specifying information 42 consisting of 3 bits and extend the bit restriction information 43 to frame data 40 in which 16 sample data of 16 bits per sample are gathered. Thereafter, the originally decoded frame data 40 is transferred to the buffer 504, and the frame data 40 stored in the buffer 504 is output as an output signal of the buffer 504 through the output terminal 508, whereby the PCM Decryption to data is complete.

1 is a block diagram showing a schematic configuration of a digital signal encoding apparatus A according to a first embodiment of the present invention. The block diagram which shows schematic structure of the digital signal encoding apparatus B which concerns on the 2nd Embodiment of this invention. 1 is a block diagram showing a schematic configuration of a digital signal decoding apparatus C that decodes a bitstream E1 generated by a digital signal encoding apparatus A according to a first embodiment of the present invention. The block diagram which shows schematic structure of the digital signal decoding apparatus D which decodes the bit stream E2 produced | generated by the digital signal encoding apparatus B which concerns on the 2nd Embodiment of this invention. The figure which shows an example of the digital signal used for description of the encoding of the 2nd Embodiment of this invention. The figure explaining the code length of the predictor identifier and prediction coefficient specific information in a bit stream. The figure which shows an example of a structure of a bit stream typically. The figure which shows the other example of a structure of a bit stream typically. The table figure which shows the average bit length at the time of performing Golomb encoding with respect to a predetermined random signal. The figure which shows an example of a digital signal. The figure which shows the appearance frequency for every amplitude in an audio | voice signal. The figure which shows the appearance frequency for every amplitude in a prediction residual signal.

Explanation of symbols

A, B: Digital signal encoding device C, D: Digital signal decoding device 201 ... Buffer 202 ... Predictor 203 ... Encoder 204 ... Code length calculator / selector 205 ... Bit stream generator 210 ... Prediction coefficient table 212 ... Subtractor 213 Memory unit 502 Bit stream decoder 503 Predictor 504 Buffer 505 Prediction coefficient forming unit 507 Prediction coefficient table 509 Memory unit 510 Bit range extension unit 511 Path switch

Claims (7)

A digital signal encoding device for reversibly encoding an input digital signal,
Sample acquisition means for acquiring the input digital signal every predetermined number of samples;
A plurality of prediction coefficients applied, and at least one prediction means for predicting a prediction signal of the current digital signal from a past digital signal by each prediction coefficient;
At least one residual signal calculation means for calculating a residual signal between the prediction signal predicted by the prediction means and the current digital signal of a predetermined number of samples actually acquired by the sample acquisition means;
The residual signal calculated by the residual signal calculating means and the prediction coefficient specifying information for specifying the prediction coefficient applied when predicting the prediction signal used for calculating the residual signal are encoded. At least one first encoding means;
Second encoding means for encoding a digital signal of a predetermined number of samples acquired by the sample acquisition means without going through the prediction means;
Encoding that selects one of the first encoding unit and the second encoding unit based on the code length of the code obtained by the first encoding unit and the second encoding unit A selection means;
A digital signal encoding apparatus comprising: Further comprising effective bit number detecting means for detecting the effective bit number of the digital signal of the predetermined number of samples acquired by the sample acquiring means,
The second encoding means comprises:
2. The digital signal encoding apparatus according to claim 1, wherein the effective bit number detected by the effective bit number detecting means and effective bit number specifying information for specifying the effective bit number are encoded.   2. The digital signal encoding apparatus according to claim 1, wherein the prediction coefficient in the prediction means includes a prediction coefficient calculated by linear prediction analysis from the digital signal of a predetermined number of samples acquired by the sample acquisition means. .   The prediction coefficient in the prediction means includes a prediction coefficient obtained by vector quantization based on a prediction coefficient calculated by linear prediction analysis from the digital signal of a predetermined number of samples acquired by the sample acquisition means. 2. The digital signal encoding device according to 1.   The digital signal encoding apparatus according to claim 1, wherein the prediction coefficient in the prediction means includes the prediction coefficient used in the first encoding means selected in the past by the encoding selection means. A digital signal encoding method for lossless encoding of an input digital signal,
A sample acquisition step of acquiring the input digital signal every predetermined number of samples;
A plurality of prediction coefficients are applied, and each prediction coefficient predicts a prediction signal of the current digital signal from a past digital signal;
At least one residual signal calculation step of calculating a residual signal between the prediction signal predicted by the prediction step and the current digital signal of a predetermined number of samples actually acquired by the sample acquisition step;
The residual signal calculated by the residual signal calculating step and the prediction coefficient specifying information for specifying the prediction coefficient applied when predicting the prediction signal used for calculating the residual signal are encoded. At least one first encoding step;
A second encoding step for encoding the digital signal of a predetermined number of samples acquired by the sample acquisition step without going through the prediction step;
Encoding that selects one of the first encoding step and the second encoding step based on the code length of the code obtained by the first encoding step and the second encoding step A selection process;
A digital signal encoding method comprising: An effective bit number detecting step of detecting an effective bit number of the digital signal of the predetermined number of samples acquired by the sample acquiring step;
The second encoding step comprises:
7. The digital signal encoding method according to claim 6, wherein the effective bit number detected by the effective bit number detection step and the effective bit number specifying information for specifying the effective bit number are encoded.

Images